{"id":"W4312516176","doi":"10.1162/tacl_a_00511","title":"Causal Inference in Natural Language Processing: Estimation, Prediction, Interpretation and Beyond","year":2022,"lang":"en","type":"article","venue":"Transactions of the Association for Computational Linguistics","topic":"Topic Modeling","field":"Computer Science","cited_by":194,"is_retracted":false,"has_abstract":true,"ca_institutions":"Columbia College","funders":"","keywords":"Causal inference; Computer science; Interpretability; Inference; Artificial intelligence; Causality (physics); Natural language processing; Robustness (evolution); Machine learning; Interpretation (philosophy); Data science; Econometrics","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003941802,0.00007093567,0.0001008973,0.0001233042,0.0003168692,0.00004702772,0.0002557437,0.00002842414,0.000002414777],"category_scores_gemma":[0.0009841025,0.0000722758,0.00004290715,0.0003348594,0.0000183859,0.0001102657,0.00003591328,0.0001756872,2.204072e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002491818,"about_ca_system_score_gemma":0.0001593505,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002313914,"about_ca_topic_score_gemma":0.00001817758,"domain_scores_codex":[0.9989764,0.00007125798,0.0003271464,0.0001689033,0.0003533844,0.0001028822],"domain_scores_gemma":[0.9986883,0.0005336072,0.0002607605,0.0001122602,0.0003871779,0.0000178934],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000008095644,0.00004575151,0.001715385,0.00002348255,0.0000133278,1.078577e-7,0.00266121,0.9693738,0.00001071988,0.01874815,0.00002380278,0.007376159],"study_design_scores_gemma":[0.0003364751,0.00002943389,0.005066474,0.00001024079,0.00001571947,0.000001365653,0.0001567182,0.974398,0.00002364614,0.0197674,0.0001321985,0.00006230678],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.005193481,0.00009538954,0.9929555,0.0004822177,0.0007656324,0.0002407088,0.00008363696,0.00004569585,0.0001377565],"genre_scores_gemma":[0.9406909,7.745049e-7,0.0589601,0.00008152298,0.00003035457,0.00004747977,0.00003173177,0.000005392746,0.0001517759],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9354974,"threshold_uncertainty_score":0.2947322,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.008751666553274558,"score_gpt":0.2654126314121237,"score_spread":0.2566609648588491,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}